Full resolution of film!

[Nikon 2]

All depends. For example, Velvia 50 resolution depends on where you mean on its repro scale. We ole Cibchrome printers skated on about a 3 stop range of it. But nowadays you've got scanners and digital manipulators who claim they can get way more juice out of the lesson by taking advantage of what lays deeper down in its density. One of he problems in that case, is that the graininess is bigger down in that pit, and the color balance is also skewed. Better to leave it alone and keep it black.

Few lenses can even resolve certain subtleties of Velvia hues anyway; and that is an independent question from lens resolution ability. But in the present day, all such distinctions might be lost to the relatively poor gamut qualities of inkjet reproduction anyway. But the reason I'm even mentioning this is that a lens regarded as superior in one respect, might be less superior in another manner.

For example, certain of the Zeiss-branded (Cosina) Nikon lenses took into account a certain look which might improve what we consider ideal color, yet at the expense of ideal resolution. With other lenses, even from the same company, it might be the opposite. They naturally try to cross-market these to both film and digital users, but you can't always ideally favor both. I'm mainly a large format shooter, and Nikon made the special M series of LF lenses for sake of higher color accuracy, using less air-glass interfaces, yet at the same time, that reduced certain other favorable characteristics like optimal image circle size.

There simply is no one shoe fits all answer. But it gets downright silly when I see people lugging around a $3000 lens on the latest 35mm digital camera, all for sake of nothing more than sharing their shots on the web! It wouldn't make any difference if they had spent 90% less.

I appreciate your understanding…!

 

[snusmumriken]

Is it true in order to get the full resolution of the highest [finest] grain film, one must have a lens capable of allowing this?

The consensus would appear to be yes. Seems obvious really.

 

[Nikon 2]

The consensus would appear to be yes. Seems obvious really.

Does Nikon have a lens capable of this…?

 

[miha]

Does Nikon have a lens capable of this…?

I guess all the lenses you have listed in your signature line are capable of such.

 

[BrianShaw]

I would have asked the question differentl; are there any Nikon lenses that, by virtue of their design, can not potentially…

 

[Nikon 2]

I would have asked the question differentl; are there any Nikon lenses that, by virtue of their design, can not potentially…

I would have asked the question differentl; are there any Nikon lenses that, by virtue of their design, can not potentially…

Interesting...!

 

[miha]

I would have asked the question differentl; are there any Nikon lenses that, by virtue of their design, can not potentially…

I can think of one or two and I'm sure there are more ...

 

[Helge]

I've posted a bit about this in the Copex Rapid Advice thread. Getting the highest resolution out of film is relatively straightforward, if not exactly easy - use a super-high-res film like CMS 20 or Copex, a very sharp lens, and a tripod. You can avoid the tripod by only shooting in bright sun (on a sunny day, Copex will let you use 1/250 or so at f/5.6) or maybe using an IS lens. But not an IS zoom, because you likely won't get the resolution you need out of it.

Now, is it worth it? Well, who knows? I can offer a bit of context, at least. Here's a shot taken on Copex at f/5.6 with a pretty cheap Rikenon P 50/1.4 on a Pentax ME Super.

View attachment 347757
This is scan composed of several stitched shots from a 24mp NEX-7. The final version is 12739x8539, coming out to an equivalent of nearly 9000dpi. I think it's around 177lp/mm? Here's a crop from near the center. Bear in mind that at least on my 27" 2k monitor (109ppi), viewing this at 100% equates to an almost 10-foot print.
View attachment 347758
Now, this is a 108-megapixel image, but it obviously isn't resolving 108 megapixels of detail. I would say it resolves about as well as a 36mp sensor, depending. Here's what it looks like reduced to 36mp:

View attachment 347759
And, because why not, a screengrab of what it looks like on my monitor previewing a 30x45 print, which is the largest standard-ratio I could get made at any typical commercial lab:
View attachment 347760
Whatever else, let's appreciate that it's fully possible to get a very clean, sharp 30x45 print out of 35mm film!

But now...let's put that side-by-side with a similar 30x45 preview from my 24mp original RX1:
View attachment 347760 View attachment 347761
The conclusion I can draw right away is that if you don't plan on printing bigger than 30x45, shooting these hi-res films is probably not worth it vs. digital, especially if you're aiming for shooting on trips. They're not very convenient to use, thanks to their low ISO and odd developing requirements, and don't (in my opinion) impart any sort of interesting character to an image the way Tri-X or HP5 do. And there are tons and tons of cheap 24mp digital cameras out there that'll give you what you need. And they shoot in color!

It also shows that no, you do not need an $8000 Leica lens to make the most of film resolution. I'm sure the Summicron APO would be marginally better than this (and would likely have improved corners), but I paid less than $100 for the camera and lens combined. Speaking of which, if you're going to be selective about what you shoot, $6 a roll for Copex makes this not a bad bargain vs. the digital option, especially in terms of buy-in cost.

Here's one more comparison set. First, again, is the 100% crop from the 108mp Copex scan, and second is the RX1 image enlarged to the same size (and de-saturated for fairness).

View attachment 347758 View attachment 347764

I hate enlarged digital images, and I'd pretty obviously choose the Copex shot here if I had to pick one of these to print at 10' wide. I would actually hang the 10' print of the film shot on my wall and have no problem with people sticking their faces right up to it! However, when we swap the 24mp RX1 for a 61mp A7RIV (this shot is from DPReview's sample gallery), it's a different story.
View attachment 347758 View attachment 347766
I suppose I might still prefer the film shot from a subjective standpoint thanks to the organic nature of the grain, but the digital one is now technically better. And, again, the original's in color. So if you've got the money for one of the modern 61mp cameras and a lens that can handle it, that's the way you should probably go. Still, at $2400 for, say, a used A7rIV with a Tamron 35/2.8, you could shoot a lot of Copex before you started losing money in the comparison!

Sorry, but not convincing.
For starters, it doesn’t appear that you emptied out the Copex frame at all.
There is a bunch of pixels evident along contrast edges, and some very soft structure of what appears to be grain. Could be the usual amalgam of noise and grain alias

Not too impressed with the 61MP example either.
Seems to be held back by optics judging by the type of softness. Also displays the typical mushy interpolated smoothness typical of digital were there wasn’t any contrasty edges to grab on to.

We really have no way of knowing for certain to which standard you grabbed any of the frames.

There are a number of caveats and pitfalls with cam scanning that it’s easy to fall into even if relatively experienced.

Also without shooting the exact same scene and with exact same same crop a comparison like this is close to useless in isolation.

Stuff like CMS 20 isn’t anymore difficult to travel with than any other film. In fact due to the low speed probably easier.

It needs a special developer to get the most out of it, that is readily available, but it works OK in other devs too.

 

[faberryman]

Why so little focus on the empirical resolution of films and the empirical resolution of lenses? If a film has higher resolution than a lens, or vice versa, doesn't the answer present itself?

 

[BrianShaw]

Why so little focus on the empirical resolution of film and the empirical resolution of lenses?

Apparently it requires too much thinking and understanding...

Talking in vague terms is so much easier...

 

[BrianShaw]

Apparently it requires too much thinking and understanding...

Talking in vague terms is so much easier...

Plus, the ultimate goal may be more to inspire conversation than to actually get answers to the question asked...

 

[Helge]

Why so little focus on the empirical resolution of films and the empirical resolution of lenses? If a film has higher resolution than a lens, or vice versa, doesn't the answer present itself?

If only it was that easy. They work together in “mysterious” ways. And what answer exactly?

Plus, the ultimate goal may be more to inspire conversation than to actually get answers to the question asked...

Which is exactly what happened.

 

[chuckroast]

I would need to look at the charts, but my guess that as film improved from say Kodak Plus X at around 125 LPM (?) to Tmax 200 at 200 LPM the later version of Nikon lens, manual and AF were likely tweaked for improved resolution. Adding to modern designs, multicoating, better and flare control will also improve acutance. , The most improvement would seen in zooms, the new ED version are much better than the 60 and 70s versions. There are likely Nikon guys with hard data. Still I think most modern lens high end lens, Nikon ED, Canon L , Leitiz, Minolta G, Pentax, LE Sigam Arts in the wide to short tele range will resolve Tmax 200. The better lens will have less distortion wide open and stopped down to F16 to 22.

I concur with this. Modern zooms are hands down better in almost every way than the old manual focus Nikkors.
It is almost certain that some of the primes have better correction. For example, my 35mm f/1.4 Ai-S is lovely stopped down but has a lot of visible coma wide open. It is reported - not confirmed by me personally - that the newer autofocus variant of this lens corrects considerably for this.

Since I don't want to own two completely different sets of lenses for film and digital - well, my bank account doesn't - I have one modern zoom for my digibody and everything else is Ai or Ai-S to be used in either medium. I just keep the limitations of these lenses in mind as I use them.

 

[Paul Howell]

As this a 35mm thread, when in College in the 60s I minored in Photojournalism, my college did not have a photography major but it did teach journalism. In one class we were discussing Leica, Context and Nikon lens. We got into the weeds of lens resolution. Our instructor finally chimed in. What he said made sense then and now. For fine detail dont use 35mm, use MF and LF, 35mm is to capuater the moment, the news. For press work, you are shooting for newspaper spread, large, half page above the fold, most much smaller, then halftone printing kills details. It's all about the moment.

 

[BrianShaw]

As this a 35mm thread, when in College in the 60s I minored in Photojournalism, my college did not have a photography major but it did teach journalism. In one class we were discussing Leica, Context and Nikon lens. We got into the weeds of lens resolution. Our instructor finally chimed in. What he said made sense then and now. For fine detail dont use 35mm, use MF and LF, 35mm is to capuater the moment, the news. For press work, you are shooting for newspaper spread, large, half page above the fold, most much smaller, then halftone printing kills details. It's all about the moment.

Well, up to a point. As discussed in many threads on this forum, 35mm film (etc) techniology has massively/gigantically/ginormously/significantly changed/improved since the 1960.s

 

[Nikon 2]

Well, up to a point. As discussed in many threads on this forum, 35mm film (etc) techniology has massively/gigantically/ginormously/significantly changed/improved since the 1960.s

It better have...!

 

[faberryman]

If only it was that easy. They work together in “mysterious” ways. And what answer exactly?

I thought the question in post #1 and the response in post #2 set up the issue nicely. Then it pretty much went off the rails, culminating in post #132 with a photo of a giant coffee cup that looks like a lens, which I guess was supposed to mean that if you drink too much coffee not even a Leica APO lens will achieve the full resolution of film because your hands will be shaking so much you can't hold the camera steady. Or something.

Is this really "mysterious" or do we just not have the film and lens resolution data to answer the question objectively?

 

[MattKing]

It better have...!

Why?
No journalist is going to care much.
Most portrait photographers won't care much.
Most people viewing the results "normally" on a monitor or screen won't care much.
Are you making billboards from your work?

 

[miha]

Or something.

To the point.

 

[Nikon 2]

I thought the question in post #1 and the response in post #2 set up the issue nicely. Then it pretty much went off the rails, culminating in post #132 with a photo of a coffee cup that looks like a lens, which I guess was supposed to mean that if you drink too much coffee not even a Leica APO lens will achieve the full resolution of film because your hands will be shaking so much you can't hold the camera steady. Or something.

 

[Leidolf]

Is it true in order to get the full resolution of the highest grain film, one must have a lens capable of allowing this? So maybe a Leica 50mm f/2 APO ASPH should be the type of lens needed to get all the resolution from film…!

Nope. save for old pre WW1 box cameras FILM is always the limiting factor. Especially colored films....
And deducing the resultant resolution is much like deducing the resultant sesistance in an electrical circuit, when adding resistors in a serial coupling :
1/N1 + 1/N2 + 1/N3 etc etc

N1 being the resolution in LPM (lines pr millimetre) of the taking lens
N2 being the resolution in LPM of the film
N3 being the resolution in LPM of the enlarger lens or the projector lens

This a rough estimate but usually holds good.
A camera lens was vastly more resolution than a film...... And the resolution of any photopaper can usually de discarded, its even better than a Leica lens in the circumstances we're discussing here...

Example Leica lens :
1/1000 + 1/100 - 1/ 450 = 119/9000 = 0,0132222 or about 1/132

Example Schneider Trionar lens
1/500 + 1/100 + 1/450 = 116/1125 = 0,0142222 or about 1/142

Both are close to the max resolution of the film.
And do make a note anything you add in the chain will degrade the image... And noone will ever be close to the max resolution of the lens in question.

Fo

 

[Anon Ymous]

Example Schneider Trionar lens
1/500 + 1/100 + 1/450 = 116/1125 = 0,0142222 or about 1/142

That's about 1/70, so 70lpmm. Generally speaking, the resolution of the system is lower than the resolution of the weakest link in the imaging chain.

 

[Paul Howell]

Is this really "mysterious" or do we just not have the film and lens resolution data to answer the question objectively?

Which film, which lens, how was the negative's resolution measured? I tested many of the M42 50mm and a few others, Konica, Miranda, Minolta MD. that I had. My test rig was not too elaborate, a very old Air Force Test Chart, low quality microscope to read the LPM, good outdoor full shade lighting, heavy tripod and cable release. I did not have a M42 body with mirror lock so all were tested at highest shutter speed for the lighting wide open, F8 and F16. I used the last of my microfiche film, developed for contrast. I was not testing for distortion ,contrast or flare as I don't the gear. All the normals lens, could resolve somewhat beyond Tmax at 200LMP. Konica 50 1.7 and Pentax 50 1.4 with radioactive element were the best, close to 500 LPM.

If you find an Air Force or other chart with LPM, use ortho film, you can test your lens. But, after doing my testing, what was the point? I shoot 400 speed films on occasional roll of Tmax 100, but Foma 400 and Tmax 400 are what I carry. Every 35mm normal lens 50 to 57mm I have will resolve Tmax 100. Well maybe not Argus C 3. Then the question is does OP want to know resolution or is really asking about sharpness or apparent acutance which is subjective.

 

[Nikon 2]

Nope. save for old pre WW1 box cameras FILM is always the limiting factor. Especially colored films....
And deducing the resultant resolution is much like deducing the resultant sesistance in an electrical circuit, when adding resistors in a serial coupling :
1/N1 + 1/N2 + 1/N3 etc etc

N1 being the resolution in LPM (lines pr millimetre) of the taking lens
N2 being the resolution in LPM of the film
N3 being the resolution in LPM of the enlarger lens or the projector lens

This a rough estimate but usually holds good.
A camera lens was vastly more resolution than a film...... And the resolution of any photopaper can usually de discarded, its even better than a Leica lens in the circumstances we're discussing here...

Example Leica lens :
1/1000 + 1/100 - 1/ 450 = 119/9000 = 0,0132222 or about 1/132

Example Schneider Trionar lens
1/500 + 1/100 + 1/450 = 116/1125 = 0,0142222 or about 1/142

Both are close to the max resolution of the film.
And do make a note anything you add in the chain will degrade the image... And noone will ever be close to the max resolution of the lens in question.

Fo

Well, that opens a can of lenses…!

 

[alanrockwood]

Many years ago, in one of the major photo magazines, I think it was Modern Photography, there was an article on achieving 100 line pairs per mm. To summarize, if one used a very good prime lens at optimal aperture, using good technique (like using a tripod) you could get very close to or even achieve 100 line pairs per mm using Panatomic X film. That was measuring resolution by the method they used at that time at that magazine.